Although I am not ready to "cut" this line, I am less confident about
the documentation of the ancestry of the Scottish kings back to Fergus
than I was at the time I wrote the Henry Project pages on them. Some
modern authors have expressed skepticism about the pedigree of Kenneth
son of Alpin (e.g., Woolf, in his recent history of the period, "From
Pictland to Alba, 789-1070"), and while I believe that this skepticism
goes too far, they make some good points. In fatc, when I wrote the
pages, I indicated that the documentation for this line in the early
eighth century was far less than ideal (see, for example, the Henry
Project page for Eochaid mac Echach). Also, Fergus himself is not
strictly historical, but can be regarded as probably reliable tradition,
and the first member of the line who can be considered as "historical"
with confidence is his grandson Gabran. For this reason and others
indicated below, I would not regard Fergus as the "winner" of the
"Longest connected well-documented genealogy contest."
In fact, the quality of genealogical documentation can vary continuously
from extremely solid (multiple independent reliable contemporary
sources, perhaps even with confirmation using DNA evidence) to just good
enough to plausible to pure guesswork to downright false, with no clear
dividing line between the various levels of confidence, and no objective
way of assigning a numerical value to the quality of documentation, for
which there can be honest differences of opinion. Thus, different
criteria for the required quality of documentation would result in
different "winners" of the contest.
Using a strict criterion, if one demands a genealogy in which every
individual in the chain appears in a well-documented historical context,
then the European "winner" would undoubtedly be the Carolingian ancestor
Arnulf, bishop of Metz (d. 640), or if you prefer, a tie between him and
Pippin the Elder (or one generation later if you demand strictly
contemporary evidence).
If this strict level of documentation is required, the only possible
competitor of which I am aware would be the Japanese imperial family,
but I don't know enough details about the documentation to say whether
or not they would qualify. From what I have read, the earliest
historically verifiable emperor of Japan was Kimmei (d. 571), but the
quality of documentation for the intervening generations between him and
the present emperor is not clear, especially in the period immediately
following Kimmei, which includes individuals who might not satisfy the
"well-documented historical context" part of this criterion. I am
inclined to believe that the quality of documentation for the early part
of the Japanese genealogy would be more comparable to the Irish case
mentioned in the next paragraph, making the Japanese line a close to the
Irish case in the less strict criterion.
To my knowledge, if one excludes lines which are due either to
deliberate fraud or gross imcompetence by a modern author, all other
claims to trace the genealogy of a currently living person to an
individual living prior to the year 500 involve either a large degree of
conjecture (e.g., the "dotted-line" pedigrees so often seen in DFA
attempts), or the use of what might be called "traditional" pedigrees.
The "connected well-documented" part of the criterion clearly excludes
the first of these, and that leaves the traditional genealogies (which
might themselves involve fraud or incompetence occurring in medieval
times). Here, it seems reasonable to use a somewhat less demanding
criterion, in which one only requires that each link be documented with
a reasonable degree of confidence, but we still run into differing
opinions about what "reasonable degree of confidence" means. In my
opinion, a long string of "generations" in a genealogy in which the
individuals are otherwise not identified in any verifiable context is
unaccepable, but a small number of generations of otherwise unidentifed
individuals is acceptable if sound arguments can be made that the
information goes back to a written source compiled not too long after
the individuals in the genealogy lived. The term "not too long" is
obviously (and necessarily) somewhat vague, but the main point is that a
simple statement that a string of names forms a reliable genealogy is
not acceptable, and some clear and convincing argument needs to be given
as to why the information should be accepted.
That being said, I have never seen any reason to believe that the
genealogy of the descendants of Confucius has an acceptable level of
documentation in the above sense. I have no doubt that there are
certain early pieces of the genealogy which mention documented
historical individuals, but the accounts that I have seen include too
many strings of a dozen-plus generations which are nothing but names,
with no apparent chronological context beyond generation-guessing, and
no corroboration from independent historical sources. If I am wrong
about this, I am willing to be convinced otherwise, but I want to see a
discussion that amounts to more than "trust me, the genealogies are
accurate."
As I have written in this newsgroup before, there is a strong case for
accepting a connected genealogy back to a number of Irish familes to
around the year 500 or perhaps a generation or two earlier, and of many
Irish families back to the seventh century. The Irish had an active
vernacular literature by the seventh century, and the amount of early
Irish genealogical material is vast, amounting to perhaps 10,000+
individuals between the years 500 and 1200. Unfortunately, much of the
material survives only in later manuscripts, and by the eighth century,
the Irish scholars were fabricating an elaborate "pseudohistory" going
back to the time of the Flood. Many figures from vague legend were
worked into the genealogies and placed in an allegedly precise
chronological framework (complete with falsified chronicles), and other
figures were invented (or in some cases duplicated) to fill in the 2000+
year gap. The result was a remarkable "history" extending for thousands
of years, the earliest part clearly fictional to all but the most
gullible, and the latest part historical, with the fictional and
historical blended in such a way that the boundary between the two is
far too often unclear. Fortunately, the falsifiers usually did not
cover their tracks that well, and careful study by scholars has revealed
much about the methods used. The most reliable of the historical
records is the annals, and it has long been recognized that the survivng
annals have a common source which existed in the early tenth century,
and was itself a compilation based on earlier annals. Careful analysis
by numerous modern scholars has revealed that the earliest contemporary
Irish annals were written in the middle of the sixth century. Such
annals were extended, interpolated, copied, recopied, combined,
recombined, etc. in numerous uncertain stages, resulting in what we have
today. Genealogies were also being compiled, extended, copied, recopied,
etc. during the same period.
When I first found these genealogies as a young enthusiast, I didn't
have the library sources to look at them in much detail, but after I got
access to more detailed material, it didn't take long for the bubble to
burst.
The early part of this "history" is a complete mess, and there are
simply too many genealogical and chronological inconsistencies to make
any reasonable sense of it, even if you assume for the sake of argument
that there is some basis to it. If one examines the chronological and
genealogical material as a whole, it holds together as a plausible whole
fairly well back to about 500 or a little earlier, shows some serious
problems by the middle of the fifth century (apparently partly due to
attempts by early Irish scholars to push back the date of St. Patrick),
and becomes an inconsistent mess by ca. 400 and earlier. This fits
quite well with the studies showing that the earliest contemporary
annals occurred ca. 550. This does not mean that these records can be
regarded as universally reliable back to ca. 500. Far from it. Later
material was sometimes interpolated into the original contemporary
framework. This is very difficult material, and each case needs to be
examined individually. Of the paternal ancestors of the famous king
Brian Boruma (d. 1014), only his father appears in the annals, and it
has long been recognized that the earlier part of his pedigree allegedly
showing his relationship to the Eoganachta has been fabricated. The
paternal ancestry of Diarmait Mac Murchada (father of Eve of Leinster)
is problematic prior to the early ninth century. However, some families
are better documented. Through her mother Mor Ua Tuathail, Eve was a
descendant of Muiredach mac Murchada (d. 760), king of Leinster, who has
one of the best documented genealogies of any individual of the early
middle ages. A genealogy of Muiredach written probably during his
lifetime names all eight of his great-grandparents and twelve of his
sixteen great-great-grandparents. Some of these individuals are
extremely obscure, but others are mentioned in the annals. Muiredach's
ancestry illustrates one of the recurring problems with early Irish
genealogies (and of similar "traditional" genealogies in general). We
can say with reasonable confidence that most of Muirdach's genealogy is
reliable back to the late sixth century, and probably for a few
generations before that, but it is seldom possible to prove the exact
generation at which it becomes mythical. It is the old "trying to
separate the wheat from the chaff" problem. While I believe that it is
highly likely that there are numerous individuals appearing in the
genealogies from the fourth (or even the third) century who actually
existed, the Irish scholars did so many cut-and-paste operations on
their depictions of the early legends that there is no way of separating
the real from the fake or the hopelessly mangled. For this early part,
even in cases where the names go back to an individual who did in fact
exist, there is no way to be sure that the pseudohistorians have put
them in their proper chronological or genealogical contexts.
Due to the fact that there are large numbers of individuals who can
document a direct male line back to early Irish kings, it is possible
that knowledge of the early Irish dynasties will improve in the future,
but I have not been impressed by the genealogical scholarship that I
have seen in some of the early papers on the subject. It is not enough
to test the y-DNA of a few individuals with a given Irish surname and
then assume that their y-DNA is that of the presumed ancestor of the
family. A good "paper trail" is required, and this is a problem for
Irish genealogies, because the slimness of records for the period
1600-1900 often makes it difficult for individuals to trace back to the
time when they might connect in to the traditional genealogies.
Some time ago, I did an ancestor table for Llywelyn ap Iorwerth of
Wales, which includes his Irish ancestors, including Muiredach mac
Murchada. It is still on the GEN-MEDIEVAL page at Rootsweb:
http://www.rootsweb.ancestry.com/~medieval/llywelyn.htm
My own website includes a detailed analysis of the early medieval kings
of Osraige, whose history becomes problematic before ca. 650.
http://sbaldw.home.mindspring.com/Ireland/Osr/lists/Osraige.htm
As for other traditional genealogies, the Welsh genealogies are, in
general, much thinner and based on later evidence. Contemporary
annalistic writing in Wales has been clearly established only for the
late eighth century and later, making confirmation of the traditional
genealogies more difficult. The main Welsh line that would rival the
Irish lines is that of the kings of Dyfed going back to Vortipor,
mentioned by his contemporary Gildas (probably early sixth century) and
also by an apparently contemporary monument stone, but the verification
of the intervening generations is slim, although one early independent
Irish source helps (see the above Llywelyn URL). Of the Anglo-Saxon
dynasties, the only one having proven modern descendants is the West
Saxon dynasty, and their genealogy is problematic. Even if one accepted
the genealogy back to Cerdic (early sixth century), despite significant
red flags in the intervening generations, the generations prior to
Cerdic are a proven fabrication. The traditional Scandinavian
genealogies aren't even in the race. There are modern lines descended
from the dynasty of the kings of Dublin going back to the late ninth
century (with a missing generation; see the Llywelyn ancestor table),
but none of the saga genealogies can be verified even that far back.
Stewart Baldwin